Epilepsy is a significant public health problem, with most of the medical, social and economic burden of disease falling on those with medically intractable epilepsy(MIE). The goal of our research is to identify, improve and disseminate technologies which treat MIE in the most cost-effective way. The primary aim of this study is to constrain previous estimates of the cost-effectiveness of epilepsy surgery derived from modeling studies. Secondary aims are 1) to identify patient/provider characteristics associated with favorable cost- effectiveness, 2) assess accuracy of patient reports of health care utilization and 3) assess the validity and responsiveness of generic, preference-based measures of health-related quality of life in this population. Subjects will be patients in a large, multi-center cohort study of the outcomes of epilepsy surgery. We will collect prospective data on health care utilization, using patient diaries and provider administrative databases. Costs will be calculated from provider accounting costs and the Medicare Fee Schedule. Benefits will be calculated by summing quality adjusted life years(QALYs), using health state utilities as quality weights. In the primary analysis, a cost- benefit difference will be defined as the difference between A) net costs and benefits accrued during the presurgical evaluation period (projected over a two year time horizon and excluding evaluation and surgical costs) and B) the net costs and benefits accrued during evaluation, surgery and the two-year follow up period. QALYs will be valued at 50,000 dollars in the reference case analysis. Modeling approaches will be used to extend the time horizon in secondary analyses. Multiple regression analysis will identify patient and provider predictors of favorable cost-benefit differences. Known-groups analysis, regression analysis, and Guyatt's responsiveness statistic will test criterion validity, convergent/discriminant validity and responsiveness. Study results will provide data to policy makers regarding the value of epilepsy surgery in treating MIE. Identifying patient and provider based characteristics associated with lower costs and better outcomes will lead to improvements in patient selection and quality of care. Identifying the most valid utilization and preference assessment methods will help future investigators make informed, data-based choices of measures for outcomes studies in patients with epilepsy and other chronic, neurologic diseases.